*The author Andres Saarniit is senior expert of Eesti Pank Macroeconomic Research Department.


One of the main economic problems of our time is that of reducing budget deficits. With the exception of the last few decades, government budget deficits in Europe have been rare. Other exceptions are the years of the war and depression. Even though Keynes recommended before World War Two that expenditures greater than revenues be used as a way to avoid cyclic variations in the economy, budget deficits only became common in the 1970ies. At this time pro-deficit policy was already losing support. Oddly, notable budget deficits only developed when they were no longer a theoretical policy goal and therefore more pragmatic reasons for this phenomenon are being investigated.

The reason usually cited is the slowdown in economic growth after the 1950ies and 1960ies as well as loosening up of the budget policy. With the decreased rate of economic growth, social insurance structures created in the 1960ies became an increasingly larger burden on society, despite the fact that the rise in inflation could conceal the increasing deficits.

A budgetary policy based on borrowing cannot be explained with social indices progress. In fact, much of the progress in European countries (the rise in life expectancy and literacy as well as the decrease in infant mortality) had been achieved by the 1960ies, that is, before debts started mounting. The social development after the 1960ies does not indicate that countries which spent more have reached a higher level of development.

For neoclassical thinkers such as Barro, the budget deficit represents a possibility of taking loans as a stopgap measure in times of crisis in order to avoid raising taxes. Tax burden, they believe, should remain consistent to avoid market repercussions. Much discussion has also centered on whether and to what extent budget deficits are related to a country's form of government.

The increase in debts servicing costs and the rise in real interest rates - possibly a byproduct of financial markets globalization and rising budget deficits of the major world powers - has forced states with large debts to strive for reducing their deficits. This is also a prerequisite for the European Monetary Union.

Estonia is a special case in the sense that although EU accession has been declared an economic policy priority for this year, deficit size as a criterium for accession has escaped our attention. In addition, the terminology we use to discuss this issue is completely different from that of international practice.


In an academic treatment, measurement of budget deficit size is not a problem. But it is a different story in applied economic policy. The balanced budget (that is, the budget with no deficit) has been one of the foundations of the economic policy of newly independent Estonia. Paradoxically, however, the international understanding of budget balancing is unknown to us, and regulation of budget deficit is not a fiscal target.

Current Estonian conventions are different from international practice (see Table 1), and we do not use terms such as overall deficit or financial deficit. Figuratively speaking, Estonia uses what could be called a ledger book system, where revenues and expenditures are always equal. Economic and political aspect, especially in the long term, is inevitably left aside.

According to international practice, the overall deficit is equal to the net volume of the loans taken by the general

government (central government including the local governments and off-budget funds) which is necessary for balancing the difference between revenues with expenditures and loans given to the non-financial system.**

**Unless a special reference is made, the terms government and deficit are used in the above meaning in this article. (Author's note).

The overall deficit is not affected by onward government lending to banks, but among the factors having an impact on deficit are included loans to structures not in the financial system. The latter are necessary in cases when companies do not possess sufficient credibility to receive loans on their own. Since loans to non-financial system structures affect the overall deficit very differently from country to country, the term financial deficit is widely used. It means the overall deficit without the non-financial-sector loans. The role of such loans can be insignificant. For example, when the Maastricht Treaty says that the budget deficit must not exceed 3% of GDP in the EU states, it is referring to the financial deficit.

In Estonia, however, the distribution of such loans in the first years after the restoration of independence was the primary reason for deficit (see Table 2). Apparently, this fact contributed to the wrong impression that the overall deficit has secondary importance in fiscal policy, because the loans to producers were "not proper public sector expenditures" from the view of national accounts. Since the Estonian government loan policy is only a few years old, the costs of servicing debts are not comparable to the size of the overall deficit (interest is expected to remain at about 0.5-0.6% of GDP). In the EU, interest is usually greater than the overall deficit and the primary deficit, i.e. overall deficit without interests is positive. In other words, everyday expenditures and investments of the government sector are covered with current revenue, with a surplus of tax revenue. In the last two years, the EU budgetary policy, reflected by a primary deficit, has been successful in terms of reducing the overall deficit. Further reduction has been hampered by the high debt burden (an average of 70% of GDP). Reduction of the debt to 60% of GDP is another EMU requirement in the Maastricht Treaty. Estonia has experienced the opposite process: expenditures and investments, not the costs of servicing earlier made debts, have increased the overall deficit.

The above factors have apparently deprioritized the budget deficit issue in the eyes of the Estonian public. Related indicators have been publicized in foreign publications, but not domestic ones. Only last year were the overall deficit and financial deficit figures published in Estonia. Even then, the first, Ministry of Finance-published nine-month deficit figure went without the attention it should have received.

The budget deficit of the Estonian government has been both positive and negative in the 1990ies, going by international measurement standards. Even though the size of the deficit has been conservative (from 1993-1996 the size of the average deficit in the EU countries was reduced from 6.2% to 4.4%), it is still larger than the optimum 1% of GDP. In addition, the fact that vernacular terminology ignores international nomenclature can hurt foreign investors' confidence, especially if the government proclaims that the budget is balanced when actually the overall deficit has been growing rapidly over the last few years.


Many states require their governments to practice balanced budget policy. In Estonia, the budget deficit will be regulated through deficit financing (only partially implemented right now). The Borrowing by the Republic of Estonia and State Guarantees for Foreign Loan Agreements Act ties borrowing to central government budget revenues. In addition, a new state budget bill and a bill of the restrictions on local government borrowing have been worked out.

Under the new budget bill, restrictions on foreign loans are extended to all loans. Both the central government budget bill and the local government budget bill set forth three limits (these of course may change, as with any draft law):
   1. the net flow of new loans cannot exceed 15% of budget revenues;
   2. loan servicing costs cannot exceed 20% of budgetary revenues (15% for local budgets);
   3. the total loan burden cannot exceed 75% of budgetary revenues.

Still, these restrictions are not sufficient to answer affirmatively the question whether Estonia intends to practice a Maastricht-style budgetary policy.

First off, these draft laws contain no clear idea of a budget deficit. In fact, some provisions both bills suggest that borrowed money will continue to be considered revenue. The Statute of Tallinn, adopted in autumn 1996, also treats loans this way.

Second, the central and local budgets are treated separately, and loans to non-budgetary funds are left to the jurisdiction of the statutes of such funds.

Third, central government and local government revenues are not clearly separated. It is not apparent whether transfers from the central to local governments are included as local government revenue or not. The most meaningful restriction is the requirement that loans can only be used to finance investments (this derives from the requirement that current revenue and expenditures are balanced).

The separate treatment of various budget levels is very notable in accounting as well: in many publications, only information on foreign loans taken or guaranteed by the central government is printed, while information on loans taken by local governments is published in their monthly reports, and information on the loan burden derived from local bond issues is available from the Central Depository for Securities. It is the local governments that have started active borrowing from domestic as well as foreign sources in recent times (see Figure 1). Since there is no unified, overall accounting system covering all levels of government and all loans, we cannot be sure whether the government's entire loan burden at the end of 1996 was 9 or 10% of GDP.

With the current tax rates, when the above bills become law, the permissible loan net flow will be about 4% of GDP, loan burden 20% and loan servicing costs 5.5% of GDP. Compared to the Maastricht requirements, these limits are quite slack. There is also the danger of a self-fulfilling prophecy: that these criteria will start to be considered an aim, for example that the loan burden should be increased to 16% (about the Latvian level) or 20% of GDP.


Deficit policy is inseparable from general expenditure-cutting (presuming that preservation of the stability of the tax burden is considered necessary).

Generally, Estonia's loan burden and level of expenditures is considered moderate. In 1994-1995, about 40% of GDP was re-distributed through the budget. Depending on tax policy (relations between the income tax-free minimums and excise rates and inflation, proliferation of local taxes, etc.), general government expenditures can be forecast for the next couple of years at 38-39% of GDP. Considering the fact that more affluent states can afford a greater tax burden, government expenditures in Estonia is not really so small.

In developed industrial countries, private enterprise is considered more tolerant of a greater tax burden because of the greater manufacturing efficiency, and thus government expenditures can exceed 50% of GDP, for example. According to this model, the ideal figure for Estonia would be less than 35% of GDP. Economies in transition generally do not fit into this model, however.

The breakdown of expenditures in Estonia is fairly unusual. As can be seen in Table 3, consumption expenditures (including salary expenditures) are large and are in line with those of more affluent states. The low level of overall expenditures is due only to the low level of subsidies (down to 0.5% of GDP) and the fact that the costs of servicing foreign loans have been small so far.

Similar to most transition economies, transition in Estonia started with a reduction in investment in the public sector. From 1992-1994, such investment from current revenue was only around 1.5% of GDP. Later that figure has exceeded 2%. By 1995 it was 4% and last year 4.6%, according to preliminary figures. This was only due to foreign loans and greater borrowing by local governments from the domestic market.

Since it has only been possible to finance investments by loans, and the costs of servicing them have still been low, there is no public pressure for greater austerity. Along with the growth of the overall deficit, we have noticed a tendency of growth of consumption expenditures of the government (see Figure 2). There is no precedent for cutting such expenditures in Estonia (it was only declared official policy at the start of 1997) and even the Riigikogu (the Parliament) tends to overestimate revenues. It is unlikely that the central government or local governments would be able to cut expenditures by 15% in a given year. There is a tendency instead to take additional loans. Interest levels can fluctuate, however. Also, the level of interest rates is important in the context of general economic growth. Estonian prices are currently approaching world market levels together with the rise in productivity. The debt burden will decrease naturally in these circumstances, as long as new debts do not accrue.

Even the inevitable rise in pension costs from an aging population does not seem to be a stimulus for greater austerity. The pay-as-you-go pension system (PAYG) that Estonia is adopting is seen as a factor increasing the budget deficit in the future. Average European states faced by increasing pension costs are already thinking of reducing government expenditures.

Estonia's situation is unusual in some regards. The ratio of the pension-age people to working-age population was approximately 38% in 1996. Most EU countries will reach this dependency ratio level only in 20-25 years (see Table 4). Thus, Estonia's pension expenditures are larger compared to many EU countries. In the EU countries pension payments ran between 4.5% of GDP (G.B.) and 11.1% (Denmark), but in Estonia they have grown from 6.6% in 1993 to 8% in 1996. Due to immigration that has affected the demographic situation and the rise in the pension age, the ratio of pensioners to the working-age people will go down to 28% by the year 2010 and reach the current level again after the year 2020. This creates the illusion that pension expenditures will become easier to cover (the replacement ratio will increase by nearly 1.4 times by the year 2010, if all other conditions remain the same) and that there is no need to focus on cutting expenditures. There is also no stimulus, or possibility for the emergence of an American-style pension system based on personal savings.


Estonia's budget has become a somewhat neglected part of economic policy. The deficit, which started growing in the first half of 1995, had reached an estimated 1.6% of GDP in 1996 (in 1995 overall deficit was 0.8%). The government's debt burden has reached about 10% of GDP. A comparison with EU indicators (in 1996 the overall deficit was predicted at 4.4% and debt burden 73.5% of GDP) may paint a rosy picture, especially when not taking into account the difference between the primary and overall deficit in the EU and Estonia. Foreign investor confidence can also be shaken by the lack of proper nomenclature in this field, especially since policy declares a balanced budget while the overall deficit has grown rapidly in the past couple of years.

The overall budget deficit should become a separate economic policy goal. Many experts feel that policy-making is even more significant here than getting legislative restrictions in place.

The new state budget bill and the bill restricting borrowing by local governments links borrowing with yearly budget revenue. The permissible net flow of new borrowing should be around 4% of GDP following the passage of these laws; loan burden should be around 20% and servicing costs about 5.5% of GDP. However, the 20% loan burden could be seen as recommendable. In such a case the loan burden increases to the level that it would have to be serviced by even more loans.

It is unlikely that governments are able to reduce expenditures by 15% in a given year, but rather more likely that they will take more loans. Interest rates can fluctuate, however, and these rates are important in light of economic growth. Estonian prices are meeting world levels with the rise in productivity and the deficit will naturally decrease, provided new debts are not taken.

Andres Saarniit

 1. Eesti majandus 1996. aasta 9 kuu jooksul. Tallinn, Rahandusministeerium, november 1996.
 2. Alesina, A., Perotti R., Fiscal Adjustments in OECD Countries: Composition and Macroeconomic Effects. IMF Working Paper 96/70.
 3. McDermott, C. John, Wescott, Robert F., Empirical Analyses of Fiscal Adjustment. IMF Working Paper 96/59.
 4. World Population Projections 1994-95, World Bank, Washington, D.C., 1994.